In-homogeneous precipitation of Iron from SC1 Solutions
- PDF / 1,635,898 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 54 Downloads / 261 Views
ABSTRACT It is found that Fe in a ppb level Fe-spiked SCI solution precipitates in-homogenous. The local high iron concentrations are difficult to rinse, create micro-roughness, and induce weak spots in capacitors. With AFM it has been shown that rings of 3-8 PLm are etched in the silicon and that the etch products (silicates) are deposited on one side just outside the ring. The capacitor yield loss induced by this SCI treatment can not be recovered fully by a subsequent SC2 or dHCL step, while a dHF/dHCI does fully recover the yield and removes the iron and silicate rim. INTRODUCTION The standard cleans I and 2 (SCI and SC2) of the RCA clean are widely used in ICmanufacturing. The SCI is a mixture of ammonia, hydrogen peroxide, and water mixed in a volume ratio of 1/1/5 and is also known as the APM. In a SC2 or HPM, the ammonia is substituted for hydrochloric acid. SC I is a cleaning solution that can effectively remove particles, but metal contamination is hard to remove. To achieve this an SC2 cleaning step is applied. It has been additionally demonstrated that to remove metal contamination, SC2 can be replaced by diluted HCI (dHCl), as the hydrogen peroxide in SC2 decomposes rapidly, thus making the addition of costly hydrogen peroxide obsolete [1]. However, immersion of hydrophobic wafers in an iron contaminated SC I solution leads to
the formation of so-called "clustered LPD's". These clustered LPD's appear parallel to the immersion direction of the wafer into the SC1 bath. This phenomenon is not observed when hydrophilic wafers are immersed in Fe-spiked SC1. The clustered LPD's have been shown to correlate with defects in capacitors. Moreover, subsequent SC2 or dHCI cleaning steps were not successful in recovering the yield loss induced during the Fe-spiked SCI step [2]. Although, it was concluded at that time that surface micro-roughness was at the origin of these defective capacitors, we will propose a model for iron precipitation on hydrophobic wafers and demonstrate its effects on the surface microroughness and the gate-oxide integrity. Additionally, a dHF/dHCI rinse step will be evaluated for recovery of the Fe-spiked SC I induced yield loss. EXPERIMENT The chemicals used for these experiments were of gigabit or better quality (M < 0.1 ppb): hydrogen peroxide (30%) and ammonia (25%) of Ashland. A standard Fe(N0 3)3 Merck (1000 wppm) solution was diluted to I w-ppm or 10 w-ppm stock solutions. With these stock solutions the APM was spiked to 0.1-10 w-ppb levels. The quartz containers and wafer carriers were cleaned with boiling diluted nitric acid (5%) for one hour and subsequently rinsed with DI-water. Monitor silicon wafers (n or p type, [100] orientation, 150 mm diameter) were FSI-B cleaned and dipped (5 minutes) in a SPM bath followed by immersion in a 0.5% HF solution. APM solutions were prepared from 5 L H20, 1 L H20 2, and 1 L ammonia. These solutions were 63 Mat. Res. Soc. Symp. Proc. Vol. 477 01997 Materials Research Society
made by heating the DI-water optional salts to 75-80°C. Subsequently
Data Loading...
